Universal CJK-Aware XLSX Handling

Problem It Addresses

When working with Excel files, the practical risk is not just calculation but deliverability: Chinese CSV exports may use GB18030 or GBK, CJK text can render as tofu boxes, and naive openpyxl save cycles can discard VBA macros, pivot tables, charts, and slicers. Formula repairs can also leave stale calcChain.xml behind. This skill treats spreadsheet editing as a verifiable engineering task rather than a blind write-back.

How It Works

It routes work around READ → ACT → VERIFY:
- READ: use xlsx_reader.py to discover sheets, macros, pivots, and other structure, then analyze with pandas without modifying the source.
- CREATE: build new workbooks from a minimal XML template, prefer Excel formulas over hardcoded numbers, then pack with xlsx_pack.py.
- EDIT / FIX: unpack the original XML, modify only the targeted cells, preserve vbaProject.bin, pivotTables/, charts/, and related parts, then repack.
- VALIDATE / REVIEW: run formula_check.py for #REF!, circular references, and similar issues, recalculate with libreoffice_recalc.py, and preview with xlsx_render.py using CJK-capable fonts.

Boundaries and Notes

It fits office scenarios with Chinese headers, full-width punctuation, RMB or Chinese date formats, and wan/yi unit display; financial color rules are optional. For visual design, complex chart reconstruction, or non-Excel data pipelines, additional tooling is needed. When processing existing files, do not assume UTF-8 CSV exports and avoid generic load/save round-trips on workbooks with macros or pivot tables.